According to this article, a traumatic brain injury turned a furniture salesman into a mathematician. (Not without side effects, but still.)

There is a bit of conventional wisdom in evolutionary biology that drastic improvements in efficacy are not available through trivial modifications (and that nontrivial modifications which are random are not improvements). This is an example of the principle that evolution is supposed to have already 'harvested' any 'low-hanging fruit'. Although I don't think much of this type of website (note the lack of external links), the story seems to be based in reality; it is thus one of the most surprising things I have ever heard. And, oddly, heartening as well---insofar as it suggests both a potential shortening of the timescale for human intelligence augmentation and the possibility that such augmentation may be relatively more accessible (than I previously thought) by comparison to computer-based artificial intelligence developments.